The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill analyzes local UI components and test files using git diff and grep. This creates a surface where the agent could ingest instructions embedded within the repository's source code or diff outputs.
Ingestion points: SKILL.md (Step 1, Step 2) defines commands that read content from the packages/ui/src and test/ directories.
Boundary markers: No specific delimiters or markers are instructed to isolate the output of the analysis commands from the agent's control instructions.
Capability inventory: The skill utilizes shell execution for project-specific commands (pnpm test:e2e), source control (git), and file searching (grep).
Sanitization: No explicit sanitization or filtering is performed on the content of the files before they are processed by the agent.
[COMMAND_EXECUTION]: The skill instructs the agent to execute various shell commands including git, grep, pnpm, and psql to manage the testing lifecycle and environment. These commands are limited to the local development environment and are appropriate for the skill's stated purpose of managing E2E tests.
[CREDENTIALS_UNSAFE]: Documentation examples contain hardcoded default credentials (e.g., PGPASSWORD=payload) and connection strings (postgres://payload:payload@localhost:5433/...) for local database access. These are identified as standard vendor-provided defaults for local development and testing environments.
[SAFE]: The use of isolated ports (PORT=3150) and in-memory databases for testing represents a security best practice for development workflows to avoid environment pollution.

ui4-convert-tests